SecondBrain

A system that combines a Chrome extension and FastAPI server for content extraction, embedding generation, and LLM-powered Q&A.

Overview

SecondBrain enables users to:

1. Extract content (text, images with OCR, video transcriptions) from webpages using a Chrome extension
2. Process and embed the content into vector space using a FastAPI server
3. Query the embedded knowledge through a beautiful Streamlit chat interface powered by:
   • Local LLM (LM-Studio)
   • Cloud-based LLM (Groq API) with multiple model options

Components

• Chrome Extension: Extracts content from webpages and sends it to the FastAPI server
• FastAPI Server: Processes content, generates embeddings, and provides query capabilities
• Streamlit Chat: Modern user interface for querying the knowledge base with LLM selection options:
  • Local LLM (using LM-Studio)
  • Remote LLM (using Groq API with multiple models)

Getting Started

Prerequisites

• Python 3.8+
• Node.js and npm (for Chrome extension development)
• Docker and Docker Compose (for containerized deployment)
• LMStudio (for local LLM processing)
• Groq API key (optional, for using Groq's cloud LLM service)

Setup Instructions

1. Clone the repository

git clone https://github.com/saumitrab/SecondBrain.git
cd SecondBrain

2. Install Python Dependencies

# Install all required Python packages
pip install -r requirements.txt

# If you're using Python 3 specifically:
pip3 install -r requirements.txt

3. Install the Chrome Extension

   • Follow the instructions in chrome-extension/README.md

4. Start the Application

   • Run the application using the provided convenience scripts:
     • On Windows: Double-click run.bat or run it from the command line
     • On macOS/Linux: Execute ./run.sh from the terminal
   • Or run directly with Python:

   python main.py
   # Or if you're using Python 3 specifically:
   python3 main.py
   

   • This starts both the FastAPI server and the Streamlit UI

5. Using Docker Compose (Recommended)

   • Run all components with a single command:

   docker compose up -d
   

LLM Options

SecondBrain now supports multiple LLM options for answering your questions:

Local LLM (LM-Studio)

• Pros: Complete privacy, no API costs, works offline
• Cons: Requires local resources, potentially slower
• Setup:
  1. Download and install LM-Studio
  2. Start the local server in LM-Studio (default port: 1234)
  3. Select "Local LLM" in the SecondBrain UI

Remote LLM (Groq API)

• Pros: Fast, powerful models, minimal local resources
• Cons: Requires internet, API key, potential cost
• Setup:
  1. Create an account on Groq
  2. Get your API key from the Groq dashboard
  3. Important: Enter your Groq API key in the SecondBrain UI (required to use cloud models)
  4. Choose from multiple models:
     • Llama-3 8B: Good balance of speed and quality
     • Llama-3 70B: Highest quality responses
     • Gemma 7B: Fast and efficient
     • Mixtral 8x7B: Strong performance with longer context

Enhanced User Interface

The Streamlit UI has been redesigned for a minimal, distraction-free experience:

• Simplified Design: Clean, minimal interface focusing only on essential elements
• Everything on One Page: All controls and chat in a single view without sidebar distractions
• Collapsible Settings: Access model selection and settings through a simple expandable panel
• Source Attribution: Clear display of information sources
• Streamlined Experience: Focused entirely on the Q&A interaction

Project Structure

/SecondBrain
├── main.py               # Main application entry point
├── requirements.txt      # Python dependencies
├── run.sh                # Convenience script for Unix-based systems
├── run.bat               # Convenience script for Windows
├── chrome-extension/     # Chrome extension for content extraction
├── fastapi-server/       # FastAPI backend server modules
├── streamlit-chat/       # Streamlit-based chat interface
└── docker-compose.yml    # Docker Compose configuration

Accessing the Application

After starting the server:

• FastAPI Server: http://localhost:8000
• Streamlit UI: http://localhost:8501 (or http://localhost:8000/ui)
• API Documentation: http://localhost:8000/docs

Troubleshooting

If you encounter a ModuleNotFoundError when running the application, make sure you've installed all required dependencies:

pip3 install -r requirements.txt

For LLM-related issues:

• Local LLM: Make sure LM-Studio is running and serving on port 1234
• Groq API: Verify your API key is correct and has not expired

License

MIT License

Contributors

• Saumitra Bhanage
• Tanmay S.
• Sandeep Kumar Shitala